MONEY: Ensemble learning for stock price movement prediction via a convolutional network with adversarial hypergraph model

Zhongtian Sun, Anoushka Harit, Alexandra I. Cristea, Jingyun Wang, Pietro Lio

Abstract: Highlights•Introducing MONEY, a novel ensemble learning framework to predict stock price movement, which can capture both group-level and pairwise relations.•The is the first study to demonstrate the effectiveness of integrating auxiliary information via GNNs before using RNNs for temporal studies.•Implementing experiments on real-world dataset and significantly outperforms the state-of-the-art with steady performance, particularly under a bear market.